Anet A8

[Martin Harris - Moderator]

I'm not sure which longer screws you are referring to on the Z belt, Robin? As I understand it, the Z axis is the one operated by the twin leadscrews...but I didn't find any problems with any of the screw lengths. There again, I didn't realise that there was a manual on the SD card and worked from a manual on the net and the videos I linked to earlier...

[Robin Etherton]

Ok, whatever it's called it's the belt which runs horizontally across the machine. The ends of the belt are folded round two screws and then zip tied. I found that when sliding the carriage to the extreme right or left the side of the belt was fouling the white bits which house the end rollers. Two longer screws enabled the belt to slide over slightly and line up properly so that the top and bottom of the belt were parallel.

Robin

[Robin Etherton]

Whilst on the subject of belts I had the other one installed as per build instructions and although the bed is moving freely the two sides of the belt were not parallel. Having found out that the build video installs the aluminium "H" plate upside down I put this right which improved matters slightly. Using an extra pair of clamps provided( they dont seem to be used elsewhere) put things right. Must be good for the motor drive.

Fortunately the internet is alive with blogs on this printer. It must be the best selling gadget out there.

Robin

[Martin Harris - Moderator]

I see what you mean now - looking down on my belt shows a slight misalignment (maybe 1/16" at the middle?) but I haven't noticed any binding at the extremities or effect on operation. It might be worth correcting though - good spot!

My Y axis H crosspiece is uppermost and the belt seems to be properly aligned.

[Robin Etherton]

There must be subtle differences in the various editions. Nearly finished now but too hot inside and its a shame to waste the good weather.

Incredible at the differences in belt alignment between various builds! Even with the h piece facing down I still needed the additional spacer to get it in line.

As long as it works when its finished!

[Martin Harris - Moderator]

Fingers crossed - sounds like you're doing a thorough job...

I've no experience of CAD but I've messed around with a free program called Tinkercad which seems quite intuitive - as an exercise (and in preparation for a useful and more accurate part for my current build) I've knocked out a Tempest(ish) exhaust stub quite easily. I used Cura for the slicing...

Still very much at the learning stage - there seems to be a minimum size that is practical - my first (very small scale) six stub manifold was a disaster but a single stub at slightly larger than model scale seems quite acceptable...

[Robin Etherton]

I have just about finished now but don't have the courage to switch on tonight.

Best sleep on it and give it one more check tomorrow.

[Martin Harris - Moderator]

Posted by Robin Etherton on 24/05/2017 15:08:00:

Incredible at the differences in belt alignment between various builds! Even with the h piece facing down I still needed the additional spacer to get it in line.

Just a thought Robin - could your X and Y steppers have been incorrectly marked? They come pre-assembled with the belt pulleys and the only obvious difference between them is that the pulleys are reversed.

Edited By Martin Harris on 26/05/2017 13:02:31

[Robin Etherton]

I don't think so but worth checking.

On another tack I switched on yesterday and although everything worked as it should concerning the set up of the nozzle and stops as soon as I attempted a print . Nothing.

Seems I was tring to print a stl file rather than gcode.?

Its all greek to me. Having another try tonight.

The anet a8 support group have been very helpful.

Robin

[john kennedy 4]

Ordered mine on 21st..Can't wait.

[Robin Etherton]

At last it is at this moment printing a test gcode file.

God knows what it is , it looks like a flat rectangular thing?

The bed is juddering a bit on the return so I will need to have another go at lining up the bearings and adjusting the belt tension.

Apparently the ingus( igus? Ignus?) bearings I bought thinking I was being smart are not as forgiving as the originals which come with the machine. They are quiet though and swmbo is quite happy having it working in the house.

Just to repeat the guys on the facebook support group are really helpful.

This is going to be a great toy, I hope.

Robin

[Simon Chaddock]

Robin

Any "juddering" does not sound right.

The print head "prints" as the bed moves in either direction so any hesitation will show in the finished print quality.

The bearings should be quiet. By far the greatest noise comes from the stepper motors which can create a noticeable "tune" when printing circles or ellipses.

[Geoff Gardiner]

It's good to see so many people getting an Anet A8 - I'm loving mine!

I have just got one of these:

It's an inductive proximity sensory for auto bed leveling. It was only £2 and is rated at 5v so should be plug and play.

For that money it would be rude not to get one!

I have printed this so that I can fit it just behind the fan:

The next job is to update the firmware with Skynet so that I can get it to work.

I have downloaded the software from their facebook site. They use an Arduino IDE emulator to perform the firmware update. Trouble is, it is for PC and I have a Mac. I can borrow a PC but I like a challenge so will try and figure out how to do it with a Mac.

[Geoff Gardiner]

I have got my firmware updated and my sensor is up and running. I have made a couple of test prints and....WOW.

It was definitely worth doing. I almost got there with the Mac. The problem I had was with the serial driver - I could only find one that I had to pay for. So, ever the skinflint, I used a PC to do the upgrade.

[MaL]

I'm curious Geoff, how does the proximity sensor make such a difference?

I would understand it on a delta printer as all three stepper motors have to turn as the head traverses the plate so the electronics can compensate for height variances 'on the fly', but on a cartesian printer the z axis only moves at layer change. also because the electronics have to calculate each and every head movement any inaccuracies in the mechanical build will create errors - which don't occur with 3 discrete axes. I assume the printer now does a mesh levelling probe before each print now? Does your z height now move between layer changes with the sensor?

Once you have levelled the plate against the x and y axis height adjustment is quite simple to achieve the required first layer height within the flatness of the plate. If your plate is more than10 microns or so out of true flat then you are going to have problems with or without a proximity sensor

I do have a delta printer, along with a couple of cartesians, the delta needs some form of height sensor (in my case it is a piezo sensor), but in my experience the cartesians, once set up (I use a dial gauge in a printed holder), are no bother getting the first layer right

[Geoff Gardiner]

Hi MaL

As part of the upgrade I changed the startup G Code so that it does a 9 point level reading across the bed before each print. All I can say is that it makes a better job of (effectively) leveling the bed than I do.

[MaL]

Trying to understand exactly what the probe is doing..

Is it making a map of 9 sample bed heights during a probe and amending Z=0 as it is printing (so the head moves up and down during each layer as well as the head moving up at the end of each layer, as per a delta machine

or

the probe just decides the average head height (allowing for x,y and z offsets) from taking 9 readings from different places on the bed during the probe and calls that Z=0

I guess it is most probably the latter. If so the probe is not helping you at all in levelling the bed, just in setting the first layer height. This is a task I usually do on the fly, if it is needed, while it is laying down a perimeter skirt to get an even 'squish' before it starts the actual print.

[MaL]

For getting the bed levelled I suggest you go get one of these - **LINK** and print yourself a holder that attaches to your x carriage. You can adjust any discrepancy in the X carriage to bed parallelism by moving the Z axis leadscrews against each other. Adjusting the Y axis will require some form of adjustment screw... I assume the A8 has one. It really is quite simple to get the bed level once you can see the dial gauge indicating which way it is sloping. Once the bed is as level as it can be (neither glass nor machine plate is perfectly flat) your probe can set the print height.

[parridav]

Just put my A8 together and running a test print. now I have to decide what upgrades are needed and what to print first.

[Robin Etherton]

Parridav

You may consider;

Glass bed

Front and rear frame brace.

Y belt tensioner

Mosfet for heating bed

Probably in that order of priority.

Robin

[Geoff Gardiner]

I have been working on a radial engine for my Puppeteer:

I haven't tried fitting it yet, so may need some more work doing to it.

[Simon Chaddock]

This is my solution to the problem of keeping the Z belt tension a 'strut' that directly connects the Z stepper motor to the belt return pulley. (A view of the underside)

It is in 3 parts.

A substantial rear box strut fixed to the Z stepper motor.

And twin front struts that also carry the belt return pulley.

The front and rear struts are not fixed together but are exactly the right length and press directly on the Perspex center cross member. In this way the distance between the stepper motor and the return pulley is held fixed regardless of the belt tension. The main printer frame carries none of the belt loads. Apart from 3 longer screws for the Z stepper motor it requires no modification to the A8.

The only draw back is it is hard to alter the belt tension but with a completely rigid support structure once set it should hopefully not need any further adjustment for a substantial period.

[Tony Bennett]

here is the 1/4 scale seven cylinder radial i printed out for my biplane.

Tony B

[Simon Chaddock]

Nothing to do with planes but having built a working single cylinder "air" engine I thought I would try to go one better and design a twin.

Similar in concept and with the same valve gear as before but with bigger diameter cylinders.

"Built up" from over 40 separate pieces everything is either hollow or kept as thin as possible. You can even see the shadow of the pistons through the cylinders.

The crankshaft is glued together from 7 components and to reduce friction runs in just two bearings.

Lead weighs are added to the crank webs to statically balance the weight of the pistons and connecting rods.

With twin double acting cylinders it is self starting but with the fit of the pistons and valve gear set for for "free running" it needs a lot of air!

Edited By Simon Chaddock on 05/06/2017 13:16:25

[parridav]

Robin, Thanks for that, I think I'm going to start with the Mosfet and soldering the connections to the bed together with the Velleman print bed surface. I think it needs a proper on/off switch so I'll print a case for the power supply that includes it and a new fan duct.